Network Engineer

Role overview Design, deploy, and validate the networks that connect large GPU clusters, covering compute fabrics, storage traffic, tenant access, and out-of-band management. You’ll shape repeatable network designs and work with architecture and deployment colleagues from planning through on-site bring-up and operational troubleshooting.

Responsibilities - Design InfiniBand NDR/XDR and Spectrum-X Ethernet fabrics, including rail-optimized topologies and NVLink domain integration. - Plan capacity for storage, tenant access, and internet or cloud connections. - Create cabling plans, optics bills of materials, and labeling standards for installation teams. - Bring up fabrics, run NCCL and network performance checks, and support acceptance sign-off. - Troubleshoot network issues during deployment and ongoing operations. - Build configuration templates and tools that provide a consistent starting point for each deployment.

Requirements - Hands-on experience designing and bringing up InfiniBand or RoCE fabrics for clusters of at least 1,000 GPUs. - Experience with NVIDIA UFM and switch operating systems such as Cumulus or SONiC. - Practical knowledge of optics, transceivers, and high-density cabling. - Ability to automate configuration and validation with Python and Ansible or comparable tools.

Nice to have - Experience with GB200 or GB300 NVL72 systems. - Data center or cloud interconnect design experience. - Experience working directly with customer ML infrastructure teams.

Benefits and work setup Full-time role based in San Francisco or remote within the United States, with travel up to 25%. Benefits include medical, dental, and vision coverage for employees and dependents, a 401(k), and unlimited paid time off.